linux-kernel - Re: [PATCH v2 6/6] device property: Remove struct property

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CAKdAkRTCV5Jk2PDtRTs7OcRoN0sHx-WFXt0i5aHDFDTQY=yaEA@mail.gmail.com>
Date:   Sat, 6 Jul 2019 10:48:55 -0700
From:   Dmitry Torokhov <dmitry.torokhov@...il.com>
To:     Heikki Krogerus <heikki.krogerus@...ux.intel.com>
Cc:     "Rafael J. Wysocki" <rjw@...ysocki.net>,
        Linus Walleij <linus.walleij@...aro.org>,
        Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
        Mika Westerberg <mika.westerberg@...ux.intel.com>,
        lkml <linux-kernel@...r.kernel.org>,
        linux-acpi <linux-acpi@...r.kernel.org>
Subject: Re: [PATCH v2 6/6] device property: Remove struct property_set

Hi Heikki,

On Fri, Nov 9, 2018 at 6:22 AM Heikki Krogerus
<heikki.krogerus@...ux.intel.com> wrote:
>
> Replacing struct property_set with the software nodes that
> were just introduced.
>
> The API and functionality for adding properties to devices
> remains the same, however, the goal is to convert the
> drivers to use the API for software nodes when the device
> has no real firmware node, and use the old API only when
> "extra" build-in properties are needed.

Why do we want to do that though?

If I am reading the code correctly, the change as it is implemented
broke some of the current users of device_add_properties() (i.e. in
i2c core when we instantiate a new i2c-client from a static board
info) because it will fail to remove the properties (as the software
node will be set up as primary and device_remove_properties explicitly
looks at the secondary) on device removal.

What is wrong with having device_add_properties() and
device_remove_properties() to continue properly handling both cases
(swnode only vs mix)?

Thanks.

-- 
Dmitry